He abruptly delayed signing the long-awaited presidential executive order on (AI), citing concerns that the new controls could slow America’s race to dominate the .
The draft decree, , envisioned a voluntary system of auditing advanced AI models from companies such as Anthropic to prevent catastrophic cyberattacks.
It called for a voluntary oversight system in which developers of advanced AI models could submit their products for review by federal agencies up to 90 days before release.
Trump’s brake because of China
Administration officials briefed major tech companies on the order’s content this week, and top AI industry executives were invited to attend a signing ceremony Thursday at the White House. But Trump told reporters he hit the brakes.
“I didn’t like some aspects of it,” the president said, adding that he was concerned the order would slow US efforts to beat China in the race for AI dominance.
The government has not specified what changes might be made to the decree or when its signing will be rescheduled.
Voluntary audits and internal reactions
Among other details, the seven-page draft emphasizes that government AI controls will be voluntary. “Nothing in this section shall be construed as authorizing the creation of a mandatory governmental licensing, pre-approval, or permit requirement for the development, publication, release, or distribution of new AI models, including pioneering models.”
Despite those assurances, Trump’s former AI chief, David Sachs, has raised concerns that the voluntary checks could one day become mandatory, a senior White House official told POLITICO on Thursday.
The draft also includes language aimed at malicious actors. It directs the attorney general to enforce the Computer Fraud and Abuse Act as well as “other applicable federal criminal laws against anyone who uses AI to gain illegal access to or cause damage to a computer without authorization, or uses AI during such illegal access in furtherance of any other crime.”
Read the executive order that Trump did not sign
